Canon XLH1 does uncompressed 1080i 4:2:2 over HD-SDI

[Alister Chapman]

Yes I did mean 133Mhz PCi-Express.

The biggest reason to go un-compressed is that you can do so much more with uncompressed video in post. Try any serious grading with HDV and you'll see a screen full of macro blocks and other artifacts, DVCPRO HD isn't that much better.

4:2:2 uncompressed HD is going to key way better than even 3:1:1 HDCAM.

The whole point of HD is quality, not just definition but noise, contrast, grain, lattitude and dynamic range. ALL of these get degraded by compression to some extent.

By staying uncompressed you can perform almost unlimited layers of effects, keying and transitions and the final render will look no different to the original footage... you can't say that about any compressed format.

I do agree that uncompressed is a bit overkill, but in many respects it's actually easier to handle. The computer dosn't have to uncompress the data to display it, there are few compatibility issues, it's just a lot of data.

I have built a single CPU dual core Pentium 4 PC with a decklink HD card and a 1.5Tb raid array that can quite comfortably cope with uncompressed HD. It didn't cost me 10's of thousands of pounds, but about £2000 (roughly $2800). That's not much more than you would spend on any current HD editing PC. Sure for many projects I'll use HDV, cineform DVCPRO HD or some other compression scheme, but the cost overhead of uncompressed is so small that is seems daft not to make the most of the uncompressed output. Hard drives are getting bigger, cheaper and faster every day, so storage isn't really a problem.

European broadcasters are starting to move up to HD, yes Digibeta will be accepted for some time to come, but why limit the quality of your end product to SD when for the same cost you can deliver in HD, expanding the market for your programmes at the same time.

I am surprised to hear you are still shooting on film for broadcast when you seem to be concered about the overheads of uncompressed HD. Especially since most HD broadcasters will not generally accept footage originated on Super 16 as it has too much grain, which in turn causes too many artifacts during transmission.

The Phillips monitor is not SDi mearly component and it's not broadcast. But at only £260 it should be OK for focus checking and double up as a location monitor for the PC.

[Frank Farago]

Alister, thank you so much for your explanation! I have not had much experience with DVCPRO HD, but I do have a fairly new Sony HDW-D2000/20 VTR deck here. So, the compression used here is only of the "3:1:1" type? Now, I was not even aware of that, and that's really not all that good indeed!

I attended the Government Video Expo in D.C. last month (Nov 30 & Dec. 1), and in one seminar, someone kindof panned the 'old' Sony HDCAM format/compression. At the same time, the speaker pronounced that the HDCAM-SR is the current "cat's miaw."

How do you feel about that? I guess HDCAM SR can be 4:2:2 or 4:4:4, at the user's option, right? And what is the codec difference between HDCAM and HDCAM-SR, plus what is the bit rate difference (like, 10-bit vs. 12-bit)? Rather confusing for someone who just wants to use the darn things. :~))

"I do agree that uncompressed is a bit overkill, but in many respects it's
actually easier to handle. The computer dosn't have to uncompress the data to display it, there are few compatibility issues, it's just a lot of data." Okay, I buy that wholehartedly. But... when all is said and done and your "Ben Hur 2" is finished spit and spat, what can you do with your uncompressed master footage?

That is, you can save and master it on a monster-arrayed SCSI drive set, surely. But how will you DELIVER it to someone else, like a corporate client or broadcast facility? Will you deliver it fully uncompressed on hard drive? If so, when anyone brodcasts it over the air, cable, or satellite, they will surely need to compress it (MPEG 4 or whatever) quite a bit, won't they?

Or, if you wan to print it to tape, you would probably go with Sony's HDCAM-SR of Panny's D5, both of them employing some "light" compression algorithms, I think. So, what I am driving at, you capture everything fully uncompressed and work in uncompressed and finish in uncompressed. But when you deliver or present the final product, it will be -- as of right now -- most difficult to have a completely uncompressed presentation of the results, I would think.

"I have built a single CPU dual core Pentium 4 PC with a decklink HD card and a 1.5Tb raid array that can quite comfortably cope with uncompressed HD."
Excellent, excellent. I was just wondering, how you came to the decision to go with a single CPU and not dual-Xeon or dual-Opteron? Running it under XP Pro? Also, what sort of graphics card(s) are you employing in the set-up? Lastly, the 1.5TB of data storage, this is i na separate enclosure, I suppose? Is it SATA or SCSI, and what exact type/specs?

Incidentally, you can hardly rent an editing suite for two days for the price you had spent on your NLE system! Congrats are in order!

With respect to "uncompressed workflow," the datacine we used was from film to 2K rez scan. So, if I am understanding this correctly, going from the "uncompressed" film negative to the 2K "compression" at datacine is already a compression of data, right? I guess if you demand and pay for the 4K (or 8K) scan, you would compress the original image less?

"European broadcasters are starting to move up to HD, yes Digibeta will be
accepted for some time to come, but why limit the quality of your end product to SD when for the same cost you can deliver in HD, expanding the market for your programmes at the same time."

I have more problems with U.S.-standard 24 fps vs. Euro-style 25 fps film negative conformity/conversion that with the D-Beta vs. HDCAM issue. Yes, even in non-HD countries you may be able to deliver on HDCAM or D5 or similar tape formats now, and then they will downconvert it to SD in PAL for broadcast. And one day soon, HBO and SHOWTME will demand HDCAM-SR or its equivalent on hard drive for broadcasting features and shows. Going the upconversion route at acquisition/editing from HDV to HDCAM via the adapter may also be a good avenue. However, I found that the German and Australian broadcasters in particular are rather picky about the original material. For example, it is almost impossible to give them a Digital Betacam master in PAL when the original master is a D-Beta NTSC which was then standard converted on the Snell & Wilcox Deft or similar.

"I am surprised to hear you are still shooting on film for broadcast when you
seem to be concered about the overheads of uncompressed HD."

These days, shooting on motion picture negative film stock is actually cheaper than doing the convoluted routine of going the digital route. In fact, I am somewhat surprised that anyone argues about how much cheaper shooting on video is. Let me expand on this, less you think I've lost my mind here. ;-)

First off, you don't have to join he SUV gas guzler crowd and end up with a 100:1 shooting ratio. In my experience as a producer, If you cannot splice together a 75-90 minute feature-length film from 15 or 20 hours of shot footage, you blew it, in my opinion. (Does not apply to 'March of the Penguins,' of course. :-))

Next, if you rent the camera, most of the time and places you can get considerably lower rates on a film camera than on a CineAlta or VariCam, at least here Stateside. In the U.K., this may be reversed.

With respect to rental of MP lenses, since it now seems that most pro video shoots also employ cinema lenses, this cost would be the same, except that with a film camera, you do not need an adapter for the cine lens, so you save that money. Also, when you rent a film camera (we did both Arri and Panavision shoots), you get a package deal with lenses and mags and batteries thrown into the deal.

Setting up and lighting a set takes a fraction of the time for a film camera than when you do a video shoot. In fact, you can be a bungling idiot as a gaffer or DP, and if you use the right camera, lens, and film stock, you should still have something rather presentable a the end.

Finally, due to colors, saturation, depth of feild, etc. I personally would much rather look at footage projected from a video that had originated on film stock than on any sort of video format or direct to HD from a video camera.

Having said all this, the involvement of film kind of ends for us with the negative, because after datacine, the negative gets stored, and that's about it. Unless you are doing a traditional 35 mm film print deal, you do not need to cut the negtive, but you can later, if the need should arise.

"Especially since most HD broadcasters will not generally accept footage originated on Super 16 as it has too much grain, which in turn causes too many artifacts during transmission."

This I never heard of. In fact, we tested not only 16 mm and S16 mm cameras and Eastman and Fujifilm stock (I personally prefer Fuji, by far), but also Super 8 stock using Eastman negative stock. The telecined result looked great and robust. In fact, we even had some second unit footage shot on Super 8 negs (using the Beaulieu 5008 camera, MOS) that was telecined and cut into the D-Beta master, w/o anyone noticing it or complaining about it.

Now, with HD, I don't think I would suggest to anyone making a full-length show on Super 8. However, I personally have not heard of any complaints about 16/S16. And according to all technical specs I have here from film manufacturers Eastman Kodak and Fujifilm, the "native" (analog) horizontal line resolution of 16 mm film stock is quite a bit HIGHER than HD's 1920 x 1080 native resolution. So, shooting on 16/S16 and then finishing onto HD would mean a DROPPING DOWN in quality. That is, it can be looked upon as a downconversion, not as an upconversion.

Why would someone complain about this, I just don't know. On the other hand, I have not heard any complaints as to someone shooting on 35 mm film and then told later that he/she should have shot on HD instead, because the result would have been more acceptable from a quality standpoint overall.

[Frank Farago]

I said earlier that "even in non-HD countries you may be able to deliver on HDCAM or D5 or similar tape formats now, and then they will downconvert it to SD in PAL for broadcast. And one day soon, HBO and SHOWTME will demand HDCAM-SR or its equivalent on hard drive for broadcasting features and shows. Going the upconversion route at acquisition/editing from HDV to HDCAM via the adapter may also be a good avenue."

I did not want to confuse or mix together some of these issues. These days, (most) network or major cable broadcasters stream their programming straight from hard drives. So... even if you give them a tape, chances are they will digitize it and send it out over air/cable/sat from HD. At least, from what I have seen and the delivery specs I had to comply with.

So... this also means that if you are crafty enough, you can bypass the not inconsiderable acquisition and maintenance cost of a Sony HDCAM HDW-series studio VTR deck. And deliver the footage on hard drives, HDCAM codec. Of course, I am ot sure how you can convert your uncomopressed HD material to compressed HDCAM coded material w/o the using a Sony HDW-2000 series deck. However, there probably are ways to do this simply and cheaply via software these days. if so, could someone please let me know what this editing/utility program(s) is/are?

Based on what I learned at the GV Expo seminar, there is a movement afoot for MASTERING film and other high-value projects to HDCAM SR and also (still) to Panny's D5. Supposedly, the top dog now is Sony's HDCAM SR format. And their SRW-5500 deck is good in that it records and plays back both HDCAM and HDCAM SR signals, so it can do dual duty. On the down side, it costs a pretty penny.

With respect to DELVERING your finished piece (other than in a wholly uncompressed format), I think the HDCAM SR will become increasingly "standard," initially with the likes of HBO, SHOWTIME, etc. The HDCAM SRW-5500 deck (with an adapter) and the HDCAM 2000-series decks can output the signal as HDCAM and also as SD, for the time being.

That is why, if the facility say in Europe has an HDCAM deck right now, they can brodcast your material as SD PAL. The VTR itself will output the HD tape signal as SD OUT.

Again, the true holy grail here would be to work in uncompressed, indeeed, but then.... DELIVER in some of the more accepted compressed formats but not on tape but on hard drive. Like in D-Beta or HDCAM. If someone knows how this can be done with relative ease, i.e. encode uncompressed material to Sony's allegedly proprietary Digital Betacam (NTSC/PAL) and/or HDCAM signal/standard, please do let me know.

One of these Sony HDCAM decks can cost a cool $50,000 (we've got ours for $44K) and then the topliner SRW-5500 with all the option borads will set you back close to a hundred K. So... any which way the deck acquisition can be bypassed would be great. I am waiting!!! Somebody!! Anybody? ;-)

While we already have an HDCAM deck, I am not looking forward to hitting the lottery just to get a Sony SRW-5500 VTR for close to $100,000 with all the extra fluff. I donl' mind doing datacine from film to hard drive, but if I can then keep everyting from that point on forward on hard drive, and not having to need to print anything to tape, I would be just as happy. Again, delivering on hard drives is okay these days (especally according to the tech specs of the Digitital Cinema Initiative). You can certainly lug around a desktop PC with the DeckLink HD-SDI interface and voila, instant HD projection and broadcast is possible worldwide. I am just not sure if the footage captured and then edited on hard drives, say as uncompressed, needs to be compressed first to D-Beta or HDCAM or HDCAM SR or D5 before it can be broadcast, or what?

[Alister Chapman]

An interseting discussion.

Figures such as 4:4:4 and 3:1:1 refer to the way the picture is sampled prior to encoding. HDCAM SR can do 10 bit 4:4:4 and is currently the highest quality tape format available, and it is extremely good and very robust. At 4:4:4 the compression ratio is 4.21:1 and at 4:2:2 the compression ratio is 2.71:1. It should be noted that DV compression is 5:1, so even HDCAM SR is compressing the data almost as much as regular consumer DV! The biggest difference with SR is that by sampling the signal 4:4:4 more of the colour information is sampled before compression than even uncompressed 4:2:2.

Regular HDCAM limits the resolution to 1440x1080 for luma and 480x1080 chroma pixels (3:1:1) and then compresses the data by approx 7:1, so as you can see much of the original 1920x1080 image is thrown away before being put to tape. The recording only has a depth of 8 bits, even though it is supposedly a 10 bit system. So HDCAM is even more compressed than SD DV and if you have ever tried to do any chroma key work with DV you’ll know why compression and sub sampling are big problems.

Much of what I do involves complex keying and CGI effects. HDCAM SR is out of the question on cost grounds. HDCAM does not have any more chroma resolution than SD Digibeta, so for super clean keys I want every last pixel of color information I can get. That’s where uncompressed can’t be beaten. I wish it was 4:4:4 but the H1 is only 4:2:2 and uncompressed 4:4:4 is too big for my computer to handle. Once I have applied my effects, done my compositiong etc. I then output my finished programme to tape, again this is another pay-off for uncompressed. As my workflow is not tied to any particular compression scheme there is no risk of codec compatibility issues. For example, if you shoot on HDV which is sampled 4:2:0 (3:2:0??) and then output to HDCAM which is sampled 3:1:1 you end up with footage that only has 3:1:0 sampling, not good at all! By using uncompressed over HD SDi I can output my finished programme to the tape format of choice, this could be D5, HDCAM or DVCPRO HD, but the key is that all the work on the footage has been done at the highest possible quality before applying compression and that makes a massive difference to the final quality. It’s very easy to output my programmes via HD SDi to the required delivery format and I only need to hire a deck for a day to output the programme. In addition if I wanted to do a 35mm print I could stick the whole project on DLT tapes and take the uncompressed data for film transfer.

[Frank Farago]

Right. What I was really wanted to find out, Alison, was if one can work with material on the had drive, and then deliver the finished material (master) on HD instead of on tape. Like, by encoding/compressing the (previously uncompressed) material to the compression schemes used by the likes of Sony's D-Beta, HDCAM, and HDCAM SR. Since I believe these are "proprietary" codecs, in this case Sony's.

Since this subject area was not mentioned by you, I take it that it is not really possible. Therefore, what you have is a workflow in which you work on your material in the computer, then print it out to tape. This has basically been the standard ops in NLE editing and CG/DVE work for the past 10-12 years. I don't see anything new here.

"At 4:4:4 the compression ratio is 4.21:1 and at 4:2:2 the compression ratio is 2.71:1." We shoulld keep in mind historically that Sony brought out the HDCAM format back in 1996 or 1997, I believe. Therefore, it is not as new as spink n' span. And it blew everyone away when it first came out.

The only film I know that was "mastered" (in addition to film neg) in 4:4:4 is like the last Lucas 'Star Wars' soap. If I had the Sony SRW-5500 deck here, I would use it in the single-channel mode (4:2:2), surely. In faxt, I would say that unless you have a multimillion dollar commercial spot or a feature with a budget north of $50 million, you do not have to go for the dual-channel mode in SR. Unless it's all CG and keying. This will change, of course, in the near future.

I agree with you that for the time being, the only quality assured VTR acquisition would be with the Sony SRW-5500 deck. I am not aware any place in the United States from where this deck can be rented, used in your own facility, and then returned the next day. Mabe, in the U.K. these are more popular and readily available. You would be lucky enoguh here to be able to access one at a post house, and then expect to pay well over $1,000 for one day's work, just for the rental of this one particular deck. It's something like everyobody wants, but very few is actually willing to buy.

"HDCAM is even more compressed than SD DV and if you have ever tried
to do any chroma key work with DV you’ll know why compression and sub sampling are big problems." Maybe I am not reading you right here, Alison, but are you perhaps stating that a 1/4-inch DV tape would have better picture information on it than a 1/2-inch Sony HDCAM tape would? This would surprising to me greatly, as over here, both HBO and SHOWTIME will show movies from HDCAM tapes. But they would not take your movie or whatever and show it if given to them on a 1/4-inch DV tape. I am almost positive about this, as I am looking at a deliver specs sheet generated by Showtime Networks (NYC), dated Sept. 2005.

Again, in the U.K. this may be topsy-turvy, as I know that especially DVCam is very popular over there.

"HDCAM does not have any more chroma resolution than SD Digibeta." That may be true, but using the previous cable/satellite broadcasters as examples, whenever they show a film or concert that states "Shown in HD" or "Presented in HD," that would be from Sony HDCAM or Panny D5, definitely not from standard-def Digital Betacam.

"uncompressed 4:4:4 is too big for my computer to handle." I see. I was given a hands-on demo of the newest (U.K. headquartered) QUANTEL editing/CG workstation a few weeks ago, and I believe that with the full options installed, that baby can give you 4:4:4 uncompressed all the way through. Have you looked into that and perhaps considered getting one?

"By using uncompressed over HD SDi I can output my finished programme to the tape format of choice." Other than by using Canon's new XL-H1 with the small, 1/3-inch sensors, how else are you going to acquire your uncompressed HD footage?

"all the work on the footage has been done at the highest possible quality before applying compression and that makes a massive difference to the final quality." Well, I am not quite sure how one would be getting the higher overall finishing quality: (1) Shooting with the 1/3-inch XL-H1 and transfering the signal uncompressd, then putting it out to tape at the end, or (2) Shooting with something like the 1/2-inch HDCAM CineAlta (not the SR) and capturing it to HDCAM tape. I believe the last 'Star Wars' film was shot with the 2/3-inch HDCAM CineAlta, but maybe it was the HDCAM SR that they used. Of course, the Canon XL-H1 was not out yet then!

"if I wanted to do a 35mm print I could stick the whole project on DLT tapes and take the uncompressed data for film transfer." With respect to data-to-film Arrilaser transfer, can't you also take the HD array itself with the Firewire or USB 2 interface to a facility? Or do you need to have it on DLT tape?

[Alister Chapman]

The only proprietary codec available to NLE’s is DVCPRO HD and I would not consider using this because in my opinion it is too highly compressed and the downsampling/downscaling defeats the object of shooting HD in the first place.

>>>This has basically been the standard ops in NLE editing and CG/DVE work for the past 10-12 years. I don't see anything new here.<<<

Yes and No.. It is the same workflow but most NLE’s use some form of compression, my workflow gives me the option to work uncompressed.

>>>We shoulld keep in mind historically that Sony brought out the HDCAM format back in 1996 or 1997, I believe. Therefore, it is not as new as spink n' span. And it blew everyone away when it first came out.<<<

I agree, but HDCAM is regarded as the HD Benchmark standard and many believe it to be as good as you can get, which it is not by some margin.

My point is that I have a work flow that thanks to companies such as Canon, Decklink and others, can rival and in some respects better HDCAM, the format that most regard as the best for normal HD programme production and delivery. The total equipment outlay is tiny compared to a comparable HDCAM or DVCPRO HD setup. Yes it has it flaws and issues, it is not a replacement for HDCAM, just a low cost alternative that produces quite remarkable results. Yes I could get the latest Quantel Box, but why? What is it that a Quantel $40k box can do that a MAC or PC with a $1k dual link (4:4:4) decklink or Kona card and a $4k raid array can’t do? If you don’t want to go the DIY route and don’t mind a bit of lossless compression you could use a Cineform Prospect box ($10k) or Wafien, but the thing is that very high quality HD is now possible for little more than SD.

Yes I could take the raid array to for transfer to film or I could take DLT tapes or DDS4 or big firewire drives or a whole host of other options.

Bottom line is that there is no reason why little old me working from a spare room in my home with just £10k ($15k) worth of gear can’t produce top quality HD programmes with stunning CGi and first class CSO and keying. Most engineers or broadcast professionals would not be able to tell whether it had been made with HDCAM or almost any other much more expensive HD format. Low cost, high quality workflows like this allow people like me to spend time and money where it counts, on a good script, good lighting, a good crew and good talent. I also look forward to seeing the Grass valley Infinity camera and the Red camera, it’s about time some of the small players gave the big boys something to think about.

Oh and by the way my name is Alister

[Frank Farago]

Well, everyone seems to have a different workflow. On "Metamorphosis," a film that was shot in September-October in Austria, Germany, and Hungary, the 35mm film was developed and then immemdiately telecined (NOT datacined) to HDCAM. From the HDCAM video masters, the film was edited. The next step is the DI, and then the negative will be cut and confirmed.

One can also take shortcuts, as with everything else in life, but why bother?

I agree that these misc. technical equipment and software costs are marginal when it comes to a project's overall budget. Like... Christopher Lambert's salary was quite a bit more than the entire post and DVE budget, incuding music and SFX combined. And this would have been so even if the video would have been telecined and archived onto HDCAM SR, no just plain old HDCAM.

Again, I see that you are holding HDCAM at a rather low regard. I have never heard anything back from anyone that the material was non-broadcastable simply because it was delivered on the overly-compressed HDCAM tape. But I know that in the U.K. the technical qaulity control guys are a lot more picky. Over there, you probably have no choice any more but have to deliver your finished show on at least an HDCAM SR tape in 4:4:4. I think this trend is coming Stateside as well.

We use the HDW-2000 deck quite extensively, and most folks coming here are happy to walk away with an HDCAM master in the hand, which they usually consider a "higher grade" format that whatever else they walked in the door with. Those who would not, well they need to go to a facility where they have a Panny D5 or Sony HDCAM SR VTR installed. These are large and cumbersome to carry and set-up, and can easily decalibrate themselves if one is not careful in moving them. I personally would not let this deck out of my sight. And it is still cheaper for a client to send it out with a technician accompanying it than to have it insured for the day or week.

With regard to the workflow you are describing, it is great. I was tryting to find out how else are you going to be getting uncompressed video into this workflow of yours. Other than from the Canon XL-H1 camera, that is. If there is no other way, then in our application, we would use something like the Canon for the sort of things I believe it was designed for: location scouting footage, audition tapes, test shoots, "making of feature" docs, and even as video assist out signal capturing. It is a great machine, and would be so even if it did not have that handly little BNC 'OUT' with the writing "HD-SDI." :~))

"What is it that a Quantel $40k box can do that a MAC or PC with a $1k dual link (4:4:4) decklink or Kona card and a $4k raid array can’t do?" In a word, plenty. Also, there is no QUANTEL "$40k box" that I know of. The one that was demod to us last month was priced at US$260,.000. The top gear from QUANTEL with the accesories goes up to GBP 200,000 in the United Kingdom, I was told.

If one could get a QUANTEL for $40K, I would get two, one in the morning, and one in the afternoon! Assuming, of course, I would win some sort of lottery grand prize first. ;-))

"use a Cineform Prospect box ($10k) or Wafien." What are these exactly, Alister? External, proprietary hard drive A-V recorders or...?

"I also look forward to seeing the Grass valley Infinity camera and the Red camera." I talked with Gras Valley on this camera, as I would be much happier getting any video camera in the 2/3-inch size category than anything else in the 1/3-inch sensor/lens mount size. Just don't try to get one from Abel Cine Video Tech in NYC. I believe I heard of the "Red," but what is that, exactly?

Also, the new Sony Blue-ray XDCAM HD promises to be good, and it is a 2/3-inch variety. On the other hand, the "lightest" compression it does 1920x180 HD is something like 35 MB/sec. I find that ridiculously low. IN other words, the image must be compressed the hell out of to get the throughput down to 35 Megs per second for a "true" HD rez. How do you figure they do this at Sonyland, Alison?

[Alister Chapman]

There is no need to be rude.. Francis.

[Alister Chapman]

For those that don't know..

Prospect HD is a PC HD capture and editing solution form Cineform (http://www.cineform.com/products/ProspectHD.htm). It can capture component or HDSDi at full raster 1920x1080i sampled 4:2:2 and compressed using the 10bit version of the CHD Cineform Codec, an excellent codec in my opinion. It can also be used for 2k with an additional add-on. It provides real-time multi layer HD editing for a fraction of the cost of most other systems. It is available as a turnkey system or you can buy the parts and software and build your own system.

The Wafien HR1, again from Cineform is a stand alone HD disk recorder. It has HDSDi inputs and outputs plus firewire and gigabit ethernet. It records 10bit 4:2:2 full raster HD using the cineform CHD Codec, price around $15k.

The Red camera has it's own thread http://www.dvinfo.net/conf/forumdisplay.php?f=110 but in essence is a camera that it's maker hopes to offer with a 35mm sensor capable of a multitude of frame sizes and rate available at low cost.

The new Sony PDW F330 XDCAM camcorder is based on half inch CCD's and uses HDV type compression but at 35Mb instead of 25Mb, it is still however 4:2:0 sampling.

HDCAM Chroma resolution is only 480 pixels per line which on a 1920 pixel raster is pretty poor, DVCPRO HD is 640. 4:2:2 uncompressed, Cineform and D5 are all 960. HDCAM luma resolution is only 1440 pixels, yet D5, Uncompressed, Wafien and prospect offer the full 1920. Yet a Wafien HR1 cost $15k and a HDCAM deck at least $55k. Thats why I think HDCAM sucks, lots more bucks for considerably less bang.

Alternate cameras with HDSDi include Sony F900, X300, Ikegami HDL10, HDL20, HDL40, Thomson Viper, ARRI D20, Luma KV3000 and many others.